determinants of capital expenditure decisions in the ... · this view, managers‟ decisions are...
TRANSCRIPT
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1238
DETERMINANTS OF CAPITAL EXPENDITURE DECISIONS
IN THE MALAYSIAN COMPANIES
Masyhuri Hamidi
University of Andalas, Indonesia
ABSTRACT
This study examines Malaysian companies‟ decisions to invest in capital expenditure in the context of Managerial and
Pecking Order Hypothesis. The main objective of this research is to determine the impact of internal cash flows, insider
ownership, and investment opportunity on the capital expenditure decisions based on the two competing theories. The
Managerial Hypothesis predicts a negative relationship between the insider ownership and capital expenditure. It also
suggests that investment opportunity does not affect capital expenditure. The Pecking Order Hypothesis, on the other hand,
proposes a positive relationship between investment opportunity and capital expenditure since shareholders‟ wealth is
maximized when managers optimize investment opportunities. However, this hypothesis predicts that the insider ownership
does not affect capital expenditure. This study focuses on manufacturing companies listed on Bursa Malaysia. This study
period covers the year 2009 until 2013, and a total of 109 companies were chosen using purposive sampling. Therefore, the 5
year study period consists of 545 data. To test the hypotheses, multiple and ordered logistic regression model were used,
where capital expenditure is categorized into five ordinal categories and the control variable is divided into four groups
which are ranked to capture the variability of the capital expenditures variable. The data was analyzed using STATA
program. The main finding of this research provide evident that the internal cash flow has positive and significant effect on
capital expenditure and thus, supports both studied hypotheses. However, insider ownership and investment opportunity
show a negative and significant effect on capital expenditure. Moreover, in the Malaysian context, the issue of conflicting
interest between managers and shareholders does exist and the impact is significant.
JEL Classifications:
Keywords: Capital expenditure, managerial style, Malaysian public listed companies, ordered logistic regression
Corresponding Author’s Email Address: [email protected]
INTRODUCTION
Capital expenditure (capex) is one of the most crucial managerial decisions whether at the institutional (macro)
or the organizational (micro) levels. At the macro level, capex affects aggregate demand and gross national
product, economic development, and business cycles (Dombusch and Fisher, 1987). At the micro level, it
influences production decisions (Nicholson, 1992) and strategic planning (Bromiley, 1986). McConnell and
Muscarelle (1985) reports that capex affects and is affected by firm performance. Several studies examine
factors which influence the capex level (e.g., Nair (1979), Berndt et. al., (1980), Larcker (1983), Fazzari dan
Athey (1987), Fazzari et. al., (1988), Waegelein (1988), Madan dan Prucha (1989), Gaver (1992), and Griner
and Gordon (1995)). In general, their work extend several earlier conceptual discussions of capex including Kuh
and Meyer (1957), Dusenberry (1958), Jorgenson (1963), Kuh (1963), Jorgenson and Siebert (1968), Grabowski
and Mueller (1972) and Elliot (1973).
One of the most widely discussed determinants of capex decisions is the Internal Cash Flow (ICF) of
the firms (e.g., Myers (1984); Myers and Majluf, (1984); and Griner and Gordon, (1995)). Managerial
Hypothesis (MH) and Pecking Order Hypothesis (POH) represent two popular theoretical explanations for the
relationships (Griner and Gordon (1995). Both theories suggest a positive effect of ICF on the level of capex.
However, POH and MH hold different explanations and thus, contrasting predictions with respect to the
relationships between Insider Ownership (I_O) and Investment Opportunities (IOP), and the level of capex.
Kim (2006) studies the impact of family ownership and capital structure on the performance of Korean
manufacturing firms and reports a significant positive association between the two variables. Maury (2006) also
finds increased performance due to family ownership in Western European firms. It was further documented that
firms with high government and institutional ownerships are generally profitable and have low leverage (Fraser
et. al., 2006). In Malaysia, the importance of family-share ownership should not be under-estimated. In addition
to government ownership which is at 11% (average), institutional ownerships stand at an average of 62% (Fraser
et. al., (2006)). This unique characteristic of firms‟ ownership in Malaysia provides one of the motivations for
this study. Furthermore, empirical results are inconclusive with respect to the association between I_O and
capex. Thus, the main objective of this study is to determine which of the two hypotheses (MH or POH)
dominates the capex decisions in Malaysia.
The remainder of the paper is organized as follows. Section two presents the literature review of capex
decisions and the proposed hypotheses. Comparison between the POH and MH is presented in section three.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1239
Section four discusses the methodology and sample data. Discussion and analysis are covered in section five
while section six summarizes the paper and provides suggestion for future research.
LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
Capital expenditure decisions and firm performance have been shown to be positively related (McConnell and
Muscarella, 1985). Sartono (2001) investigates the impact of ICF and I_O on capex for companies listed on
Bursa Efek Jakarta (BEJ) using a sample of 223 companies (1994 to 1998). He reports ICF to be positively
related to capex. However, in the case of I_O, the effect on capex is not significant. This finding implies that
POH prevails in the Indonesian context. A study by Marchica and Mura (2007) using a sample of UK non-
financial listed firms (1991 to 2001) examine how alternative cash-holding policies influence the firm‟s ability
to invest. They conclude that a cash-rich firms show a strong positive impact on capex while those persistently
display low cash policy invest less in capex. Masulis et. al. (2007) conclude that the scaled change in capex has a
significantly positive effect on excess stock returns, indicating that on average, capital investments add to
shareholders value.
Pecking Order and Managerial Hypotheses
In explaining the influence of ICF on capex decision, POH assumes the superiority of shareholders‟ wealth
concept. Therefore, the level of insider ownership (I_O) is insignificant in managerial capex decisions. POH
assumes that agency problems do not influence capex decision making process (Griner and Gordon 1995;
Sartono 2001). Since shareholders‟ wealth is the main managerial priority, the level of CE will increase with the
increase in IOP. This is further strengthened in the absence of conflicts between managers and owners. It is
assumed that beneficial investments to the owners will also benefit managers through financial rewards and
career advancement (Aggarwal and Samwick, 2003). In addition, according to Griner and Gordon (1995),
managers are motivated to utilise ICF rather than external funds due to the information asymmetry. Since
managers are more informed than owners regarding both ICF and IOP, the former are in a better position to over
or under invest at the expense of limited information of the owners (Stulz, 1990),
Contrary to POH, MH emphasizes the significance of managers‟ ownership and conflict of interest
between managers and stockholders in capex decision, termed as agency cost (Jansen and Meckling, 1976). In
this view, managers‟ decisions are influenced by self interests and profit maximization (Scott, 2003). Capex
entails financing risk which may jeopardize firm‟s survival and affects the portion of managerial ownerships. It
follows that large managerial stockholdings tend to result in under-investment using internal cash flow. Thus,
this theory predicts a negative relationship between I_O and capex. In addition, MH argues that capex decision is
not determined by the availability of IOP but the personal interest of managers.
Griner and Gordon (1995) suggest three main reasons why managers with small stocks ownership
prefer to finance capex using ICF rather than external financing: (i) ICF carries low risk since firms are not
committed to pay interest and thus, low bankruptcy risk; (ii) using ICF implies manager‟s performance is
reliable since shareholders are willing to sacrifice dividend from retained earnings; and (iii) further investment
of ICF signals the perception of a long term investment opportunity and firm survival. Thus, managers would be
loyal to the firm and owners would be interested in the survivorship of the firm.
The above discussion reflects the importance of ICF in deciding the capex level. However, these
relationships need further examination to understand the interplay of managerial relationship, especially between
the managers and the owners. To date, studies of this nature are relatively scant especially in developing
economies (see for example, Fazzari and Athey, 1987; and Fazzari et. al., 1988)
Internal Cash Flow (ICF) and Capital Expenditure (CE)
Elliot (1973) performs a cross-section time series of 184 manufacturing and non-manufacturing companies. He
concludes that capex is affected by liquidity, which in turn is determined by ICF. Fazzari and Athey (1987) and
Fazzari, et al., (1988) provide further evidence that liquidity significantly strengthens the effect of ICF on capex.
Similarly, Fazzary and Athey (1987) and Fazzary, et. al.,(1988) report that ICF is an important factor in capex
decisions.
POH’s explanations on the use of ICF for capex financing is based on the argument that managers and
current stockholders do not have conflicting interests (e.g., Myers (1994) and Myers and Majluf (1984)). Rather,
asymmetric information is the main motivation for managers‟ preference to use ICF in financing capex.
Potential investors believe that managers possess and use information not available to outsiders in attempting to
maximize the current stockholder‟s welfare. Thus, using ICF to finance capex signals a positive effect on the
wealth of current shareholders.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1240
Contrary to POH, the basis for MH’s explanation is the managers-shareholders‟ conflict of interest.
Marris (1964) studies ICF and capex as focal points for this conflict and argues that managers prefer to retain
and re-invest a large portion of earnings that is in the best interest of shareholders. Jansen (1986) states that the
agency conflict increases with excess cash flow since managers attempt to increase their authority by over
investing in non-core activities. These activities may not be parallel with the shareholders‟ interests.
Meanwhile, other researchers argued that although agency conflict can be reduced, it can not be eliminated by
the market mechanism (Jansen and Menkling, 1976; Fama, 1980; Jensen and Ruback, 1983; Hart, 1983; Jensen
1986).
More recently, a study by Dalbor & Jiang (2013), using a sample of U.S restaurant industry (2002 to
2012) examine how free cash flow influence the firm‟s capital expenditure. They study found that free cash flow
and size were positive determinants of capex, whereas the economic recession was a negative determinant of
capex.
Based on the explanation above, both POH and MH propose that ICF is an important determinant of
capex, whereby the two variables are positively related. Accordingly, this lead to our first hypothesis (in
alternative form):
H1: Internal cash flow (ICF) has a positive effect on the level of firms‟ capital expenditure (capex).
Insider Ownership (I_O) and Capital Expenditures (capex)
Even though previous empirical research has investigated the effect of I_O on many issues, none has
documented the relationship between I_O and capex. For example, Haugen and Senbet (1981) study the effect of
I_O in options trading. Morck et al. (1988) look at the correlation between I_O and financial performance, and
report a significant non-linear effect. Further, Wu and Wang (2005) conclude that I_O positively affects firm
value. The empirical findings of Jensen et al. (1992) suggest that there exists interdependency among I_O, debt
level, and dividend policy. Walkling and Long (1984), Benston (1985), Agrawal and Mendelker (1987), and
Sicherman and Pettway (1987) have also provided evidence that I_O influences the levels and characteristics of
mergers and acquisitions.
A more recent study by Kim (2006) involving Korean manufacturing firms (1991 to 1998) investigates
the impact of family ownership and capital structure on firms‟ performance. His findings suggest a positive
association between the two variables. Maury (2006) also finds increase performance due to family ownership in
Western European firms. In particular, where families hold at least one of the top two positions, profitability is
improved. This suggests that family ownership lowers the classical agency problem between managers and
owners. In Malaysia, where the government and institutional ownerships are relatively high, the ownership
impact on business decisions may be underestimated. Fraser et. al., (2006) documents that firms with high
government and institutional ownerships are generally profitable and have high leverage. POH assumes no
conflicting interest between managers and the current stockholders and thus, there is no effect between Capex
and I_O. In contrast, MH argues that managers increase their self-interest in deciding the capex level. Low
managerial ownership may provide incentive for them to take risk in committing to high level capex. In other
words, increase I_O is expected to reduce managerial over-investment in capex. Accordingly, MH predicts a
reverse relationship between Capex and I_O. In line with MH argument, our second hypothesis is:
H2: Insider ownership (I_O) has a negative impact on the level of firms‟ capital expenditure (capex).
Investment opportunity (IOP) and Capital Expenditure (Capex)
Following Myers (1977), IOP is defined as a combination of real assets (assets in place) and future investment
options. According to Graver and Gaver (1992), future investment option is implied by projects supported by
research and development as well as the company‟s relative ability to determine viable investment opportunities
(in comparison to others) within its industry. Myers (1977) suggests that company‟s ability is negatively
correlated with firm value which in turn depends on future IOP. Furthermore, managers tend to reach a debt
ratio target to asset from IOP. A high asset proportion implies a high debt ratio and these findings support Smith
and Watts (1992).
Gaver (1992) argues that in determining the investment opportunity sets, companies adopt a long-term
view to balance management incentive and stockholder‟s interest. In general, the literature shows that capex is
influenced by policies and compensation plans that are incentive-based, designed to align the interests of
managers and stockholders. Wu and Wang (2005) report a strong positive relationship between announcement
effects and IOP but fail to uncover the effect for issuing firms with high IOP. According to Brailsford and Yeoh
(2004), the growth opportunities are significantly important to explain the market reaction to capital expenditure
announcements.
Jung et. al., (1995) conclude that managers of growth-oriented companies prefer to increase capital
through equity financing since it allows firms to increase capital without commitment to pay interest as in the
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1241
case of debt financing. They argue that companies with profitable IOP tend to issue equity while those with
poor IOP prefer to issue debt. In addition, they report that stock price reacts favorably to equities issued by
companies with good IOP. According to Chung et. al. (1998), share price reactions to a firm‟s capex decision
reflects the market‟s assessment of the firm‟s quality of IOP.
On the contrary, Donaldson (1961) as cited in Brigham and Gapenski (1996) summarizes that firms
prefer to utilize internal rather than external financing (from retained earnings and non-cash depreciation). Those
with ICF greater than capex requirement will invest the excess ICF in marketable securities, increase the payout
ratio, or buy the treasury stocks. Whenever the ICF is insufficient to finance new projects, it will be reflected by
decreasing the size of marketable securities portfolio, followed by selling bonds, convertible bond, and common
stock as the last resort. In Hong Kong, Fan and So (2004) conclude that the pecking order principle is practiced
to maintain target debt-equity mix. Internal equity was ranked first, followed by bank debt and new common
equity. In the US, Graham et. al., (1992), Kester et. al., (1994), Wivattanakantang (1999), and Arsiraphongphisit
et al., 2000) provide support for this conclusion.
MH argues that managers tend to over-invest if the future IOP is better than the current condition
without considering the stockholders‟ welfare. According to this view, there is no impact of IOP to company‟s
capex. On the contrary, the POH predicts that in this condition, managers will increase the level of capex to
improve stockholders‟ welfare. Therefore, capex is positively related to IOP. This argument leads to our third
hypothesis:
H3: Investment opportunities (IOP) have a positive impact on the level of firms‟ capital expenditure (capex).
RESEARCH METHODOLOGY
Data, Samples and Measurements
This research uses secondary data drawn from manufacturing companies listed on Bursa Malaysia. Data are
acquired from published company reports covering a period of 5 years (2009 - 2013). The study focuses
primarily on manufacturing companies since they are more likely to invest heavily in property, plant, and
equipment compare to those in service industries. The data are selected using purposive sampling method.
Following Emory & Cooper (1995 p.228), the following criteria were used to select the sample companies: (i)
legally registered as public companies with Bursa Malaysia during the study period; (ii) having insider
ownerships (i.e., the CEO and the board of directors are listed as shareholders); and (iii) having complete
financial reports published by Bursa Malaysia during the 5-year period.
A total of 157 manufacturing companies was obtained and based on the above criteria, 48 companies
were excluded since their dates of incorporation was later than the year 2009. The final sample of 109
companies represents various sectors as presented in Table 1.
TABLE 1: COMPANIES REPRESENTATION BY SECTOR
Sector Number %
Food
Petroleum Refineries
Rubber
Plastic
Wood
Iron and Steel
Furniture and Fixtures
Paper
Fabricated Metal
Machinery and Electrical Machinery
Equipment
Other Manufacturing
Unspecified
1
4
4
5
14
11
11
2
11
4
10
19
13
0.91
3.67
3.67
4.58
12.84
10.09
10.09
1.83
10.09
3.67
9.17
17.43
11.93
Total 109 100%
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1242
Variables and Measurements
Capital Expenditure (capex)
Capex as explained by Griner and Gordon (1995) represents the amount of fund disbursed by management to
acquire property, plant and equipment. Capex can also be described as the amount paid for additional assets to
support firm development (Sartono, 2001). As defined by Griner and Gordon (1995) and Sartono (2001), capex
might be stated as the difference between total fixed assets in the current period and total fixed assets in the
previous period. The mathematical form can be stated as follows:
Capex it = TFA t - TFA t-1
Where:
Capex = Capital Expenditure of company (i) for period (t);
TFAt = Total Fixed Asset at time (t); and
TFAt-1 = Total Fixed Asset at time (t-1).
Capex is categorized into 5 ordinal categories. In ordered logit model, observed response categories of
capital expenditure yi( ) are tied to the latent variable *
iy by a measurement model that divides *
iy
into 5
ordinal categories according to the following estimates of cut-off-points:
Y =
0 = "negative capex" if yi* = 0
1 = "positive first 25%" if 0 < yi* £ 56.32
2 = "positive between 25% and 50%" if 56.32 < yi* £152.63
3 = "positive between 50% to 75%" if 152.63 < yi* £ 481.98
4 = "positive above 75%" if 481.98 < yi*
ì
í
ïïï
î
ïïï
ü
ý
ïïï
þ
ïïï
The categorization of capex is based on a simple arithmetic procedure. Companies with negative capex
are categorized as group zero1. On the other hand, firms with positive capex values are divided into four
different categories using the median value as the cut-off point. Companies having capex value below and above
the median value is further divided into two different categories with “25 percent and below” and “75 percent
and above”, respectively.
The above categorization separates companies into more homogeneous groupings. This allows for
better comparison in terms of firms‟ capacity and growth performances among the different categories. The
classifications simply follow an ordinal categorization of firm performance. It assumes that, the lower (higher)
ordered the firm, the lower (higher) its utilization (performance).
Internal Cash Flow (ICF)
Free cash flow has been defined as the cash flow remaining after firms‟ have invested in all positive NPV
projects‟ (Lang et al., 1991, p. 319). The liquidity measure used in this study is described as Internal Cash Flow
(ICF) rather than free cash flow, although its operational definition is identical. In this study, following Lehn
and Poulsen (1989) and Lang et. al. (1991), ICF is defined as:
ICF = INC – TAX – INTEXP – PFDIV – COMDIV
Where:
INC = operating income before depreciation;
TAX = total income taxes;
INTEXP = gross interests expense on short and long-term debt;
PFDIV = total amount of preferred dividend requirement on cumulative
preferred stock and dividends paid on noncumulative preferred stock.
1 Almost 54 percent of the total usable capex observations recorded negative values. According to accounting
and finance approach (Griner and Gordon, 1995 and Sartono, 2001), a negative value of capex is not an
indication of loss to a company but only reduces its total fixed asset holding. Therefore, the decision to regroup a
negative capex into lowest ordinal category is simply to avoid the elimination of such observations from the
analysis.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1243
COMDIV = Total value of dividends declared on common stock
Insider Ownership (I_O)
I_O is the percentage of shares owned by managers and executive boards of company (i) during period (t) for
each observation period. It is mathematically formulated as follows:
I_Oit = D & C SHRSit
TOTSHRSit
Where:
TOTSHRSit = Total value of shares issued; and
D & C SHRSit = Shares owned by managers and executives of company (i)
during period (t)
(Chen and Steiner, 1999; Crutchley & Hansen, 1989; Wiwattanakantang,
1999).
According to the MH, I_O is expected to reduce the tendency of managers to over-invest in capex by
forcing them to support more of the financial consequences of their actions. Thus, an inverse association
between Capex and I_O would support the MH. POH, however, believes that there is no conflict of interest
between managers and the current shareholders and thus, no association is predicted between Capex and I_O.
The POH argument is taken as the null hypothesis and analyzed against the alternative hypothesis of an inverse
association as implied by the MH.
Investment Opportunity (IOP)
IOP is a combination of real assets (assets in place) and future investment option (Myers, 1977). The measure
used is the book value of gross property, plant, and equipment (PPE) ratio with the book value of the asset
(BVA) (Sami et. al., 1999).
IOP(t+1) = PPEt / BVAt
Where: IOP = Investment Opportunity;
PPEt = book value of property, plant and equipment in year (t) and
BVAt = book value of total assets in year t.
Interaction Effects
The Managerial Hypothesis predicts a positive interaction coefficient between ICF, I_O and IOP and in this
study, the effect is estimated by the following:
ICF(it).IOP(t +1)
I _O(it)
é
ëê
ù
ûú
Total Sales
According to Griner and Gordon (1995), sales is normally used to control for firm size and measured by total
revenues generated by company (i) during period (t). Haller and Murphy (2012) also find that firm size is one
main determinant of Capex. This variable also acts to isolate the impact of ICF, I_O, and IOP on Capex and
thus, reduce the minimum rate of bias (Gordon and Griner, 1995, Sartono 2001 and Myers and Majluf 1984).
Without the isolation process, the correlations between Capex and ICF may not be recognized, because Capex
and ICF may be positively correlated with the company‟s size, as well as IOP. Furthermore, the correlations
between Capex and I_O could be unexpected since the existence of the I_O in large companies is usually small.
In this study, total annual sales are used as a proxy to control for firm size. Therefore, sales are also
grouped into 4 different categories based on simple arithmetic procedure. All groups are equally divided so that
the effects of sales become easily identified. The cut-points are estimated as follow:
D=
1 if total sales £ 25%
2 if 25%<total sales £ 50%
3 if 50%< total sales £ 75%
4 if total sales >75%
ì
í
ïï
î
ïï
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1244
Group “1” represents the first 25 percent of the sales distribution and the second 25 percent of the data
is grouped as “2”. Group “3” falls between 50 to 75 percent of the distribution and finally, the upper 75 percent
of sales is categorized as Group “4”. Each group is identified as an ordinal category. It is assumed that, the
higher (lower) the order of the group, the higher (lower) the total annual sales.
Model Development
The ordinary regression model assumes that the distances between categories of responses are equal. In order to
avoid this basic assumption, the Ordered Logit Model (OLM) is utilized. The OLM or proportional odds models
have been widely used to analyze a latent variable model (i.e., responses dependent variable). This model
assumes that the categories of an ordinal response can be ranked, but the distances between the categories are
unknown. The ordered logistic regression model with multiple independent variables is expressed as:
y* = Xtb +et et ~ NID 0,1( )
The vector of Xt variables could be a combination of interval, ratio and categorical variables. In the
ordered logistic model observed response categories (yi) are tied to the latent variable (y
i
*
) by a measurement
model that divides yi
*
into J ordinal categories so that J-1 cut-points
are estimated. The unobserved cut-points
(threshold value) are depicted as below.
In the latent variable model, the dependent variable y* could be divided by some thresholds parameters
(i.e., limited, known, number of values), which usually must be estimated (i.e., ), and it is
essential that .
In general, the latent variable model with multiple independent variables is express as:
ln Yj '( ) = ln
pjx( )
1- pjx( )
æ
è ç ç
ö
ø ÷ ÷ =g
j+ b
1X
1+ b
2X
2+…+ b
nXn( )
Where parameters pj and are those to be estimated.
However, since the ordered logistic model estimates one equation over all levels of the dependent
variable, a concern is whether a one-equation model is valid or a more flexible model is required. A
parameterization of each of the parallel equation when a constant is present in the model is assumed
homogenous across other individual parallel model.
The final OLR model with total annual sales as a dummy variable is as depicted below:
ln Yj '( ) = ln
pjx( )
1- pjx( )
æ
è ç ç
ö
ø ÷ ÷ =g
j+ b
1X
1+ b
2X
2+ b
3X
3+ b
4X
4+dD
X 5( )
Where, the coefficients for each variable are as follows:
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1245
gj Cut - point (intercept)
b1 ICF
b2 I_O
b3 IOP
b4 interaction ICF * I_O * IOP
d if D
1 if total sales £ 25%
2 if 25% < total sales £ 50%
3 if 50% < total sales £ 75%
4 if total sales > 75%
ì
í
ï ï
î
ï ï
In the proportional odds model, the covariates have the same effect on the odds as the response variable
has at any dividing point. Different values of the covariates are regarded as shifting the response distribution to
the right (or left) without changing its spread or shape. In the proportional odds model, the cumulative logistic
model the effect of covariates on odds of response below or equal to the cut-point m in the latent variable. First,
the odds that an outcome is less than or equal to m, versus being greater than m, given x is defined as follow:
The log of the odds is defined as follows:
where m is a cut-point of m, so that
P y =m | X( ) = F tm
- Xb( )-F tm-1
- Xb( ) and
P y £m | X( ) = F tm
- Xb( ) The estimation of J-1 cut-points and the intercept for each individual parallel equation is achieved by equating
the intercept to zero.
Model specification
Measuring goodness of fit in the OLM is one that needs to be cautiously interpreted. Since the logistic
regression does not have an equivalent of the R-squared value in OLS regression, the fitted model in this study is
evaluated using (i) Akaike information criterion, AIC; (ii) The Bayesian information criterion, BIC; and (iii)
McFadden‟s R2
.
According to (Long and Freese, 2001), there is no convincing evidence that the selection of a model
which maximizes the value of a given measure necessarily results in a model that is optimal in any sense other
than the model having a larger (or smaller value) of that measure. However, it is still helpful to examine any
differences in their level of goodness of fit, and hence provide some guidelines in deciding the appropriateness
of the model. Brant (1990) test is employed to test the individuality of each parallel equation. Whenever a
significant p-value is obtained, then the parallel regression assumption is violated, and a more robust technique
is required.
Results and Discussion
This research focuses on the capital expenditure (capex) decisions by Malaysian manufacturing companies listed
on the Malaysian Bourse. Four independent variables are examined, namely, internal cash flows (ICF), insider
ownership (I_O), investment opportunities (IOP), interaction terms between ICF, I_O and IOP and total annual
sales being treated as the control variable. Capex is categorized into 5 ordinal categories while total sales are
ranked into 4 groups. According to Griner and Gordon (1995), using sales as the control variable captures the
variability of the interaction terms between ICF, I_O and IOP.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1246
TABLE 4.1: RESULTS OF ORDERED LOGISTIC REGRESSION WITH ODD-RATIOS
Form Level
Odd-ratio Level
lnp j x( )
1- p j x( )
æ
èçç
ö
ø÷÷ Coef Z-value Coef Z-value
g1 -0.1764 -0.1764
g 2 0.4491 0.4491
g3 1.1173 1.1173
g 4 2.1225 2.1225
b1 0.0009 3.09*** 1.0009 3.09***
b2 -0.0277 -2.17** 0.9726 -2.17**
b3 -0.9226 -1.79* 0.3975 -1.79*
b4 0.00003 1.58 1.0000 1.58
d2 0.4292 1.83* 1.5360 1.83*
d3 0.5287 2.21** 1.6967 2.21**
d4 0.6913 2.34** 1.9963 2.34**
LR Chi-sq (7) 57.30***
McFadden R-sq 0.0413
AIC 1353.692
BIC 1399.396
Brant (1990) test 23.44 (0.321)
Notes:
***,**,* denotes a significant value at 1%, 5% and 10%, respectively.
Values in parenthesis denote p-value.
Brant test: test conducted to identify the parallel regression equation for each of the individual
model. A significant test signifies that parallel regression assumption is violated.
McFadden R-square should be interpreted cautiously, since in limited dependent variables, the
measurement of goodness of fit is incomparable with the OLS.
Table 4.1 demonstrates that a unit increase in ICF ( b1) results in the ordered log-odds of a higher capex
group to increase by 0.0009 unit (significant at 1% level) and thus, supports the first hypothesis. However, the
proportional odds ratio for ICF ( b1) is 1.0009 times higher for high capex group as compared to other capex
group, ceteris paribus. The proposed sign of ICF is consistent with both theory (i.e., POH and MH) and also
provide support for Griner and Gordon (1995), and Sartono (2001).
b2 , the coefficient for Insider ownership (I_O), shows a reverse relationship with capex (significant at
5 percent level). This result suggests that a unit increase in b2 , the ordered log-odds of being in a higher capex
category decreases by 0.0277 units, ceteris paribus. In terms of proportional odds ratio, it is found that an
increase in b2 reduces the effect of higher capex group compared to others by 0.9726 times. The significance of
I_O towards capex is assumed to follow the managerial hypothesis and consistent with Scott (2003). This
implies that conflict of interest between managers and shareholders do exist, and provide the incentives for over-
investment on the part of the managers.
This study hypothesizes that investment opportunity (IOP) is positively related to capex. Rather, the
findings show that the IOP coefficient is negatively correlated and significant at 10 percent level. This particular
result demonstrates that companies in Malaysia tend to behave according to the argument of the managerial
hypothesis. In other words, managers attempt to capitalize on the investment opportunities more for personal
wealth maximization rather than for shareholders‟ value (Mansor & Hamidi, 2008). The proportional odds ratio
indicates that a unit increase in b3 would lower the higher capex group utilization by 0.3975 times in
comparison to the other groups.
Table 4.1 also shows that as predicted by the managerial hypothesis, the interaction terms between ICF,
I_O and IOP are positive but insignificant. Compared to other groups, the effect of the interaction terms
increases the proportional odds ratio of higher capex group nearly one time. This finding is similar and
consistent to study conducted by Fazzari and Athey (1987), Fazzari et. al. (1988), and Griner and Gordon
(1995).
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1247
At 10 percent level, the sales variable is statistically significant and the level improves as sales progress
to a higher than the basic group. In short, sales within group “2” affect higher capex level at 0.4292 units
compared to group sales “1”. Similarly the effect of sales of group “2” towards higher capex group is recorded at
1.5 times higher compared to others. Moreover, the proportional odds ratio increases to about 1.7 times for sales
group “3”, and almost 2 times higher for the highest sales group. Overall, annual sales and capex are positively
correlated. This result thus, supports Griner and Gordon (1995), and Myers and Majluf (1984).
The consistency of each of the parallel regression is tested using the procedure proposed by Brant
(1990). According to Brant (1990), if each of the coefficients is found to be similar for each single equation,
then the OLR assumption is violated. As reported in Table 4.1, the chi-square value is 23.44 (not significant) and
therefore, confirms that the coefficients across the four parallel individual regressions were homogenous.
CONCLUSION
This research extends the previous work on the capital expenditure (capex) decisions by Malaysian
manufacturing companies (Mansor and Hamidi, 2008). Based of the gap in expectations between the arguments
of the Pecking Order Hypothesis (POH) and the Managerial Hypothesis (MH), four main hypotheses are
proposed. Four independent variables are examined, namely, internal cash flows (ICF), insider ownership (I_O),
investment opportunities (IOP) and sales as the control variable. Unlike the earlier study which used the OLS
method, this research applies the Ordered Logistic Regression (OLR) model. The independent variable (capex)
is categorized and ranked into five (5) ordinal categories. The control variable is divided into four (4) groups,
which are also ranked ordinarily to capture the variability of the independent variables. This study also provides
mixed support for the two hypotheses and reveals interesting characteristics with regards to the managerial style
among the Malaysian PLCs.
As predicted by both POH and MH, the results show positive and significant impact of ICF on higher
Capex group. Both theories, however, predict different outcome with respect to the relationship between I_O
and IOP. The POH predicts no effect of I_O on Capex while the MH argues for a negative relationship between
the two. In this study, I_O is found to have a significant negative impact on higher Capex group (at 5 percent)
and thus, supports the MH. With respect to IOP, the POH suggests that it is positively related to Capex since
managers attempt to maximize shareholders‟ wealth. However, the MH holds that Capex is not affected by IOP.
Contrary to both the MH and POH, this study documents a significant negative effect of IOP on higher Capex
group.
As predicted, the results suggest that ICF and capex are positively related and thus, provide further
support for the importance of ICF in capex. In contrast to the argument of POH, there is no evidence that the use
of ICF to finance capex is due to information asymmetry. Rather, such action is motivated by managerial
intention to optimize personal benefits. The conflict of interest between managers and shareholders is detected
from the negative (significant at 5%) coefficient for I_O. This finding is consistent with Kim (2006) and
supports the MH. It reflects that in Malaysia, even in publicly owned companies, the family-effect may still
dominate. To a certain extent, the family-ownership effect influences the directors‟ capex decisions and helps to
reduce the significance of agency cost in capital expenditure decisions. Also, unlike the case of Indonesia
(Sartono, 2001), the study provides mixed results as to the dominance of POH and MH in Malaysia. More
specifically, the results partially support the MH and not fully neglect the POH, providing evidence of the
significance of agency conflict between managers and shareholders in Capex decisions.
REFERENCES
Aggarwal, RK & Samwick, AA 2003, „Why do managers diversify their firms? Agency Costs Reconsidered‟,
Journal of Finance, Vol. LVIII, no. 1, pp. 71-118.
Agrawal, A & Mandelker GN 1987, „Managerial incentives and corporate investment and financial decisions‟,
Journal of Finance, vol. 42 (September), pp. 823-837.
Arsiraphongphisit, O, Kester, GW & Skully, MT 2000, „Financial policies and practices of listed firms in
Thailand: capital structure, capital budgeting, cost of capital and dividends‟. Journal of Business Administration,
vol. 88 (October-December), pp. 72-93.
Benston, G 1985, „The self-serving management hypothesis: some evidence‟. Journal of Accounting and
Economics, vol. 7 (April), pp. 67-84.
Berndt, ER, Fuss M & Waverman, L 1980, „Empirical analysis of dynamic adjustment models of the demand for
energy in US manufacturing industries, 1947-74‟. Final Report (Electric Power Research Institute, Palo Alto).
Brailsford, TJ & Yeoh, D 2004, „Agency problems and capital expenditure announcements‟, Journal of
Business, 77(2), pp. 223-256.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1248
Brant, R 1990, „Assessing proportionality in the proportional odds model for ordinal logistic regression‟,
Biometrics, 46(4), pp. 1171-1178.
Brigham, EF & Gapenski, LC 1996, „Fundamental of financial management, fifth edition, The Dryden Press.
Bromiley, P 1986, „Corporate capital investment: a behavioral approach. Cambridge University Press, London.
Chung, KH, Wright, P & Charoenwong, C 1998, „Investment opportunity and market reaction to capital
expenditure decisions‟. Journal of Banking & Finance, vol. 22, pp. 41-60.
Cooper, DR & Emory, CW 1995, Business Research Methods, fifth edition (Richard D. IRWIN, Inc)
Crutchley, CE, & Hansen, RS 1989, „A test of the agency theory of managerial ownership, corporate leverage
and corporate dividends‟, Financial Management, vol. 18 (winter), pp. 36-46.
Dalbor, M & Jiang, L 2013, „Determinants of capital expenditures in the U.S. restaurant industry‟, The Journal
of Hospitality Financial Management, vol. 21, pp. 77 – 86.
Dornbusch, R & Fischer, S 1987, Macroeconomic, fourth edition McGraw-Hill, New York.
Dusenberry, JS 1958, Business Cycle and Economic Growth McGraw-Hill, New York.
Elliot, JW 1973, „Theories of corporate investment behavior revisited‟, American Economic Review, vol. 63
(March), pp. 195-207.
Fama, E 1980, „Agency problems and the theory of the firm‟, Journal of Political Economy, vol. 88 (April), pp.
288-307.
Fan, DKK & So, RW 2004, „What managers think of capital structure: the evidence from Hong Kong‟, Journal
of Asian Economics, vol. 15, pp. 817-830.
Fazzari, SM & Athey, MJ 1987, „Asymmetric information, financing constrains, and investment‟, Review of
Economics and Statistics, vol. 69 (August), pp. 481-487.
Fazzari, SM, Hubbard, RG & Petersen, BC 1988, „Financing constrains and corporate investment‟, Brookings
papers on Economic Activity (I), pp. 141-206.
Fraser, DR, Zhang, H & Derashid, C 2006, „Capital structure and political patronage: The case of Malaysia‟,
Journal of Banking & Finance, vol. 30, pp. 1291-1308.
Gaver, JJ 1992, „Incentive effects and managerial compensation contracts‟, Journal of Accounting, Auditing, and
Finance, vol. 7 (spring), pp. 137-156.
Gaver, JJ & Gaver KM 1992, „Additional evidence on the association between the investment opportunity set
and corporate financing, dividend and compensation policies‟, Journal of Accounting and Economics, vol. 16,
pp. 125-160.
Grabowski, HG & Mueller, DC 1972, „Managerial and stockholder welfare models of firm expenditures‟, The
Review of Economics and Statistics, vol. 54 (February) pp. 9-24.
Graham, B, Bromson, G, Ma, A & Pak, SL 1992, „The capital structure puzzle‟, Securities Journal, June, pp.
20-23.
Griner, EH & Gordon, LA 1995, „Internal cash flow, insider ownership, and capital expenditure,‟ Journal of
Business Finance and Accounting, Vol. 22 (March), pp. 179-199.
Haugen, P & Senbet, L 1981, „Resolving the agency problem of external capital through options‟, Journal of
Finance, Vol. 36 (June), pp. 629-647.
Jensen, GR, Solberg, DP, & Zorn, TS 1992, „Simultaneous determination of insider ownership, debt, and
dividend policies,‟ Journal of Financial an Quantitative Analysis, vol. 27 (June), pp. 247-263.
Jensen, MC 1986, „Agency costs of free cash flow, corporate finance, and takeovers,‟ American Economic
Review, vol. 76 (May), pp. 323-329.
Jensen, MC & Meckling, WH 1976, „Theory of the firm: managerial behavior, agency costs and ownership
structure,‟ Journal of Financial Economics, vol. 3 (October), pp. 305-360.
Jensen, MC & Ruback, RS 1983, „The market of corporate control: The scientific evidence‟, Journal of
Financial Economics, vol. 11 (April), pp. 5-50.
Jorgenson, DW 1963, „Capital theory and investment behavior‟, American Economic Review, vol.53 (May), pp.
247-259.
Jorgenson, DW & Siebert, CD 1968, „A comparison of alternative theories of corporate investment behavior‟,
American Economic review, vol. 58 (September), pp. 681-712.
Kester, GW, Chang, RP & Tsui, KC 1994, „Corporate financing policy in the pacific-Basin: Hong Kong and
Singapore‟, Financial Practice and Education, Spring/Summer, pp. 117-127.
Kim, E 2006, „The Impact of Family Ownership and Capital Structures on Productivity Performance of Korea
Manufacturing Firms: Corporate Governance and the “Chaebol Problem‟, Journal of The Japanese and
International Economies, vol. 20, pp. 209-233.
Kuh, E and Meyer, JR 1957, The Investment Decision, Harvard University Press, Cambridge.
Kuh, E 1963, Capital Stock Growth: Microeconomic Approach, North-Holland, Amsterdam.
Larcker, DF 1983, „The association between performance plan adoption and corporate capital investment‟,
Journal of Accounting and Economics, Vol. 5 (April), pp. 3-30.
Proceedings of the Australasian Conference on Business and Social Sciences 2015, Sydney
(in partnership with The Journal of Developing Areas)
ISBN 978-0-9925622-1-2
1249
Lang, LHP, Stulz, R and Walking, RA 1991, „A test of the free cash flow hypothesis: The case of bidder
returns‟, Journal of Financial Economics, vol. 29 (June), pp. 315-335.
Lehn, K and Poulsen, A 1989, „Free cash flow and stockholder gains in going private transactions‟, The Journal
of Finance, Vol. XLIV, no. 3, pp. 771-787.
Long, JS & Jeremy Freese 2001, „Predicted Probability for Count Model‟, The Stata Journal, vol. 1(1), pp. 51-
57
Madan, DB and Prucha, IR 1989, „A note on the estimation of non-symmetric dynamic factor demand models‟,
Journal of Econometrics, vol. 42 (October), pp. 275-283.
Mansor, N and Hamidi, M 2008, „Capital expenditure decisions of Malaysian listed companies‟, The Malaysian
Finance Association Proceeding. 5th
-6th
June 2008, Kucing, Sarawak, Malaysia.
Marchica, MT and Mura, R 2007, Cash holding policy and ability to invest: How do firms determine their
capital expenditures? http://papers.ssrn.com/sol3/results.cfm
Marris, R 1964, The Economic Theory of Managerial Capitalism, The Free Press, Glencoe, IL.
Maury, B 2006, „Family ownership and firm performance: empirical evidence from Western European
corporations‟, Journal of Corporate Finance, vol. 12, pp. 321-341.
McConnell, JJ and Muscarella, CJ 1985, „Corporate capital expenditures decisions and the market value of the
firm‟, Journal of Financial Economics, vol. 14 (September), pp. 399-422.
Morck, RA, Shleifer, A, and Vishny, RW 1988, „Management ownership and market valuation‟, Journal of
Financial Economics, vol. 20 (January/March), pp. 293-315.
Myers, SC 1977, „Determinants of corporate borrowing‟, Journal of Financial Economics, no. 5, pp. 147-175.
Myers, SC 1984, „The capital structure puzzle‟, Journal of Finance, vol. 39 (July), pp. 575-592.
Myers, SC and Majluf, NS 1984, „Corporate financing and investment decisions when firms have information
that investors do not have‟, Journal of Financial Economics, vol. 39 (June), pp. 187-221.
Nair, RD 1979, „Economic analyses and accounting techniques: an empirical study‟, Journal of Accounting
Research, vol. 17 (spring), pp. 225-242.
Nicholson, W 1992, Microeconomic Theory: Basic Principles and extensions, fifth edition, The Dryden Press,
Hinsdale, IL.
Rose, C 2005, „Managerial ownership and firm performance in listed Danish firm: in search of the missing link‟,
European Management Journal, vol. 23, No. 5, pp. 542-553.
Sami, Heibatollah, SM, Simon Ho, and CK Kevin Lam 1999, Association Between the Investment Opportunity
Set and Corporate Finance, Dividend Leasing, and Compensation Policies: Some Evidence from an Emerging
Market, Working Paper, Presented at Program MSi-Economics Faculty Gadjah Mada University, dated
Augustus 2nd
1999.
Sartono, A 2001, „Pengaruh aliran kas internal dan kepemilikan manajer dalam perusahaan terhadap
pembelanjaan modal: managerial hypothesis atau pecking order hypothesis?, Jurnal Ekonomi dan Bisnis
Indonesia, vol. 16, No. 1 (Januari), pp. 54-63.
Scott, WR 2003, Financial accounting theory. 3rd
edition, Toronto: Prentice Hall.
Sicherman, NW and Pettway RH 1987, „Acquisition of divested assets and shareholder wealth‟, Journal of
Finance, vol. 42 (December), pp. 1261-1273.
Smith, CWJ and Watts, RL 1992, „The investment opportunity set and corporate financing, dividend, and
compensation policies‟, Journal of Financial Economics, vol. 32, pp. 263-292.
Stulz, R 1990, „Managerial discretion and optimal financing policies‟, Journal of Financial Economics, vol. 26,
pp. 3-27.
Tandelilin, E 2001, Analisis investasi dan manajemen portofolio. Edisi Pertama, BPFE Yogyakarta.
Waegelein, JF 1988, „The association between the adoption of short-term bonus plans and corporate
expenditures‟, Journal of Accounting and Public Policy, vol. 7 (spring), pp. 43-63.
Wivattanakantang, Y 1999, „An empirical study on the determinants of the capital structure of Thai firms‟,
Pacific Basin Finance Journal, vol. 7, pp. 371-403.
Wu, X and Wang, Z 2005, „Equity financing in a Myers-Majluf framework with private benefits of control‟,
Journal of Corporate Finance, vol. 11, pp. 915-945.